Rethinking grounding-zone basal drag for improved projections of Antarctic ice loss

Ice-stream grounding zones are critical parts of the Antarctic Ice Sheet and we must be able to model how they retreat with confidence. This is because for most of Antarctica, the grounding-zone location determines the flux of ice to the ocean. As such, future grounding-zone locations (as the ice-sheet edge retreats inland) dictate the pattern and rate of ice loss and the associated sea-level rise. Ice flow across the grounding zone is opposed by a combination of ice-shelf buttressing if lateral drag or pinning points are present, and by friction between the ice and its bed - basal drag. While the loss of buttressing and ocean-forced melting of ice shelves are popular research foci, far less attention is paid to resistance at the bed. Here, we revisit the treatment of grounding zones in models and argue for a new approach to basal drag: evolving bed conditions result in highly variable drag that is not captured by existing sliding laws. We begin by reviewing the complex basal processes operating in grounding zones over different timescales, such as tides, evolving subglacial hydrology, and sedimentation. Then, to illustrate the significance of the basal drag term, we run retreat scenarios in the UFEMISM model using an accepted basal sliding law that only varies in its treatment of subglacial hydrology (via effective pressure). Appreciable differences in retreat patterns motivate us to consider how modelling grounding zones may be improved. To reduce uncertainty in ice-loss projections, in tandem with refining ice-shelf melt parameterisations, careful observations from grounding zones and physically-based basal drag parameterisations are required.